From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from cuda.sgi.com (cuda1.sgi.com [192.48.157.11]) by oss.sgi.com (8.14.3/8.14.3/SuSE Linux 0.8) with ESMTP id oAPAE48d254908 for ; Thu, 25 Nov 2010 04:14:04 -0600 Received: from mail.internode.on.net (localhost [127.0.0.1]) by cuda.sgi.com (Spam Firewall) with ESMTP id A47C013EC3BD for ; Thu, 25 Nov 2010 02:15:41 -0800 (PST) Received: from mail.internode.on.net (bld-mail17.adl2.internode.on.net [150.101.137.102]) by cuda.sgi.com with ESMTP id 9LVNxtyblP2KR7O4 for ; Thu, 25 Nov 2010 02:15:41 -0800 (PST) Date: Thu, 25 Nov 2010 21:15:37 +1100 From: Dave Chinner Subject: Re: Verify filesystem is aligned to stripes Message-ID: <20101125101537.GD12187@dastard> References: <4CED5BFC.8000906@shiftmail.org> <20101125054607.GM13830@dastard> <4CEE0995.9030900@hardwarefreak.com> MIME-Version: 1.0 Content-Disposition: inline In-Reply-To: <4CEE0995.9030900@hardwarefreak.com> List-Id: XFS Filesystem from SGI List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Sender: xfs-bounces@oss.sgi.com Errors-To: xfs-bounces@oss.sgi.com To: Stan Hoeppner Cc: xfs@oss.sgi.com On Thu, Nov 25, 2010 at 01:00:37AM -0600, Stan Hoeppner wrote: > Dave Chinner put forth on 11/24/2010 11:46 PM: > > > Because writes for workloads like this are never full stripe writes. > > Hence reads must be done to pullin the rest of the stripe before the > > new parity can be calculated. This RMW cycle for small IOs has > > always been the pain point for stripe based parity protection. If > > you are doing lots of small IOs, RAID1 is your friend. > > Do you really mean RAID1 here Dave, or RAID10? If RAID1, please > elaborate a bit. RAID10 is just a convenient way of saying "striped mirrors" or "mirrored stripes". Fundamentally they are still using RAID1 for redundancy - a mirror of two devices. A device could be a single drive or a stripe of drives. > RAID1 traditionally has equal read performance to a > single device, and half the write performance of a single device. A good RAID1 implementation typically has the read performance of two devices (i.e. it can read from both legs simultaneously) and the write performance of a single device. Parity based RAID is only fast for large write IOs or small IOs that are close enough together that a stripe cache can coalesce them into large writes. If this can't be acheived, parity based raid will be no faster than a _single drive_ for writes because all drives will be involved in RMW cycles. Indeed, I've seen RAID5 luns be saturated at only 50 iops because every IO required a RMW cycle, while an equivalent number of drives using RAID1 of RAID0 stripes did 1,000 iops... Cheers, Dave. -- Dave Chinner david@fromorbit.com _______________________________________________ xfs mailing list xfs@oss.sgi.com http://oss.sgi.com/mailman/listinfo/xfs